RDP 2016-05: Trade Invoicing Currency and First-stage Exchange Rate Pass-through 6. Conclusion

We assess how exchange rate changes affect import prices. For imports invoiced in foreign currencies (roughly two-thirds of imports) the pass-through of exchange rate changes to import prices is immediate and close to one-for-one at horizons up to two years, consistent with the law of one price. For the roughly one-third of imports invoiced in Australian dollars, we find that pass-through of exchange rate changes to import prices is only about 14 per cent after two years. Consistent with US evidence from Gopinath et al (2010), these results indicate sizeable and persistent deviations from the law of one price for Australian dollar-invoiced goods.

Our findings do not indicate that invoice currency causes this low pass-through of exchange rate changes to import prices. Rather, our findings are likely to reflect foreign exporters with a low desired degree of pass-through choosing to invoice in Australian dollars. Doing so minimises pass-through for the duration over which prices are fixed. Thus, the invoice currency serves as a sufficient statistic for firms' desired degree of pass-through over a period of at least two years. Explaining why firms would want a low degree of long-run pass-through amounts to finding sources of real rigidities – factors that attenuate the response of price changes to nominal shocks. Our sense of the literature is that, although there are a number of hypotheses, the size of the deviation remains something of a puzzle.

One implication of our results is that Australian dollar-invoicing reduces the sensitivity of importers' costs to exchange rate changes, and this may reduce the response of consumer prices to exchange rate changes. Local currency pricing implies that exchange rate changes have small effects on the relative price of domestic- and foreign-produced goods, dampening expenditure switching effects. This observation has potentially important implications for monetary policy. If local currency (Australian dollar) invoicing were to become more prevalent, our results suggest that the expenditure switching channel of monetary policy may be somewhat weakened.

To the extent that prices are sticky in their currency of invoice, our findings imply that an invoice-share-weighted exchange rate index should be sufficient for modelling aggregate exchange rate price pass-through to import prices. However, we find that changes in the trade-weighted index are also relevant, possibly because there is a low degree of nominal rigidity for some foreign currency-invoiced trade.

Another important implication of our results is that aggregate import price models imposing the law of one price are likely to be unsuitable, because the speed of pass-through is extremely slow for Australian dollar-invoiced imports. Finally, our results suggest that changes in the exchange rate might have long-lived effects on the goods terms of trade. In particular, to the extent that exchange rate changes affect import prices less than one-for-one (as we find), and if exchange rate pass-through for export prices is near complete, an exchange rate depreciation might raise the terms of trade.